T1234: A distortion-matched structural scan solution to misregistration of high resolution fMRI data

Chung (Kenny) Kan,Ruediger Stirnberg,Marcela Montequin,Omer Faruk Gulban,A Tyler Morgan,Peter Bandettini,Laurentius Huber
DOI: https://doi.org/10.1101/2024.09.19.613939
2024-09-20
Abstract:Purpose: High-resolution fMRI at 7T is challenged by suboptimal alignment quality between functional data and structural scans. This study aims to develop a rapid acquisition method that provides distortion-matched, artifact-mitigated structural reference data. Methods: We introduce an efficient sequence protocol termed T1234, which offers adjustable distortions. This approach involves a T1-weighted 2-inversion 3D-EPI sequence with four spatial encoding directions optimized for high-resolution fMRI. A forward Bloch model was used for T1 quantification and protocol optimization. Twenty participants were scanned at 7T using both structural and functional protocols to evaluate the utility of T1234. Results: Results from two protocols are presented. A fast distortion-free protocol reliably produced whole-brain segmentations at 0.8mm isotropic resolution within 3:00-3:40 minutes. It demonstrates robustness across sessions, participants, and three different 7T SIEMENS scanners. For a protocol with geometric distortions that matched functional data, T1234 facilitates layer-specific fMRI signal analysis with enhanced laminar precision. Conclusion: This structural mapping approach enables precise registration with fMRI data. T1234 has been successfully implemented, validated, and tested, and is now available to users at our center and at over 50 centers worldwide.
Neuroscience
What problem does this paper attempt to address?